Thematic description of factors linked with extended-spectrum beta-lactamase-producing Enterobacteriaceae in humans

Background Extended-spectrum beta-lactamase (ESBL)-producing Enterobacteriaceae are associated with serious antimicrobial-resistant infections in Canadians. Humans are exposed to ESBL-producing Enterobacteriaceae through many interconnected pathways. To better protect Canadians, it is important to generate an understanding of which sources and activities contribute most to ESBL exposure and infection pathways in Canada. Objective The aims of this scoping review were to thematically describe factors potentially associated with ESBL-producing Enterobacteriaceae colonization, carriage and/or infection in humans from countries with a very high human development index and describe the study characteristics. Methods Four databases (PubMed, CAB Direct, Web of Science, EBSCOhost) were searched to retrieve potentially relevant studies. Articles were screened for inclusion, and factors were identified, grouped thematically and described. Results The review identified 381 relevant articles. Factors were grouped into 13 themes: antimicrobial use, animals, comorbidities and symptoms, community, demographics, diet and substance use, health care, household, occupation, prior ESBL colonization/carriage/infection, residential care, travel, and other. The most common themes reported were demographics, health care, antibiotic use and comorbidities and symptoms. Most articles reported factors in hospital settings (86%) and evaluated factors for ESBL-producing Enterobacteriaceae infections (52%). Conclusion This scoping review provided valuable information about which factor themes have been well described (e.g., health care) and which have been explored less frequently (e.g., diet or animal contact). Themes identified spanned human, animal and environmental contexts and settings, supporting the need for a diversity of perspectives and a multisectoral approach to mitigating exposure to antimicrobial resistance.


Introduction
Antimicrobial resistance (AMR) is a real and growing public health threat (1).Infections caused by extended-spectrum beta-lactamase (ESBL)-producing bacteria are a major concern because beta-lactam antibiotics are commonly used to treat a variety of infections, and some classes, such as third-generation cephalosporins and monobactams are listed as critically important for use in human medicine by the World Health Organization (2,3).Further, infections with ESBL-producing bacteria are associated with increased likelihood of severe illness and mortality and can result in treatment failures, which can lead to increased hospital-stay duration and hospital costs (4,5).
In 2018, it was reported that approximately one in four bacterial infections in Canada were resistant to first-line antibiotics, which led directly to approximately 14,000 deaths (5).Additionally, AMR has been reported to lead to negative socio-economic outcomes, including increased healthcare costs, loss of productivity, increased inequality and decreased trust in the government and public health agencies (5,6).Therefore, AMR consequences are far-reaching and have widespread implications to humans, animals and society.
Extended-spectrum beta-lactamase-producing Enterobacteriaceae have also been detected in companion animals, livestock, wildlife, water, soil, vegetables, meat and seafood, all of which can be possible sources of exposure for humans (11,14,15).Because of the variety of exposure pathways, a One Health approach that considers the interconnections between humans, animals, and their shared environments is required to cover the full scope of this growing public health threat (16,17).
Past systematic reviews have explored factors associated with ESBL-producing Enterobacteriaceae colonization and infections (13,(18)(19)(20)(21)(22)(23)(24)(25)(26)(27); however, systematic reviews are intentionally narrow in scope, providing knowledge on specific research questions.This project aimed to describe the breadth of factors previously reported to be associated with ESBL-producing Enterobacteriaceae in Canada or similar countries.This information could be used to inform various parallel projects within the Public Health Agency of Canada, such as the Integrated Assessment Model of Antimicrobial Resistance (iAM.AMR) project (15,28), and to help better understand Canadians' exposure to antimicrobial-resistant bacteria.Therefore, the objectives of this scoping review were 1) to thematically describe factors potentially associated with ESBL-producing Enterobacteriaceae colonization, carriage and/ or infection in humans from countries with a very high human development index, and 2) to describe the study characteristics.

Methods
Below the methods are described in brief.For a full description of the methodology, refer to Goltz et al. (29).

Protocol registration
An a priori protocol of this scoping review is available online.This review followed the methodological framework described by Arksey and O'Malley (30), and the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) extension for Scoping Reviews (31).

Search strategy
Search terms and databases searched are described in the protocol document.Four databases (PubMed, CAB Direct, Web of Science and EBSCOhost) were searched through the University of Guelph McLaughlin Library to retrieve potentially relevant articles.The search string for this review was adapted from Murphy et al. (32), with consultation from co-authors, in addition to a University of Guelph librarian.All databases were filtered to only include articles published in English.The initial search was completed in August 2020 and updated in August 2021.

Eligibility criteria
To meet the inclusion criteria, articles needed to be primary research, be from countries similar to Canada with a very high human development index (33), be written in English, and contain quantifying associations between factors and ESBLproducing Enterobacteriaceae colonization, carriage and/ or infection in humans.No articles were excluded based on publication year, study population characteristics (e.g., age, sex or health status) or study setting (e.g., household or hospital).These inclusion criteria were selected because of the Canadian focus of this article, and therefore aimed to identify articles with Canadian and similar populations.Further, only English articles were included due to available language resources.Relevant systematic reviews and meta-analyses were excluded but their reference lists were used to identify additional articles that were not captured by the search.

Selection of articles
The DistillerAI tool feature was used to screen titles/abstracts.The DistillerAI tool was trained by two reviewers using 226 articles.Once trained, all titles/abstracts were screened by the DistillerAI tool and a human reviewer.Title/abstract screening conflicts were resolved by a third human reviewer.Articles included based on title/abstract had the full text screened by two SCOPING REVIEW reviewers and conflicts were resolved through discussion by the two reviewers.

Data charting
Following full text review, relevant data were charted using DistillerSR by a single reviewer.Data extracted included: publication year, study design, country region (based on World Health Organization regions) (34), data collection method (primary, e.g., questionnaire or interview; secondary, e.g., database or medical charts), sample setting (e.g., hospital), outbreak episode, age of participants, microorganisms evaluated, type of colonization, carriage, or infection evaluated and factor themes (n=13).A factor was defined as a measured observation (e.g., penicillin use) that was investigated for its relationship with ESBL-producing Enterobacteriaceae (32,35).Individual factors were grouped into 13 themes created through an iterative process informed by previous work (15).Themes were 1) antimicrobial use (i.e., antibacterial, antiviral, antifungal), 2) animals (i.e., contact with animals), 3) comorbidities and symptoms (i.e., conditions or presenting symptoms), 4) community (i.e., factors that occur in the community), 5) demographics, 6) food and consumption, 7) health care (i.e., factors that occur in a hospital setting or are related to receiving health care), 8) household (i.e., factors that occur at the home), 9) occupation (i.e., factors related to employment), 10) prior ESBL colonization/carriage/infection, 11) residential care (i.e., factors that occur in a residential setting such as a nursing home), 12) travel (i.e., factors related to international travel) and 13) other factors (i.e., factors that did not fall into a previously defined theme).If a factor belonged to more than one theme (e.g., patient took antibiotics while on vacation), it was recorded in all relevant themes (e.g., antimicrobial use and travel).

Study characteristics
A summary description of the articles is reported in Table 2. Of the 381 articles included, 378 were observational study designs, and three were experimental.Most of the studies (n=235) were conducted in European Region countries, including six multinational studies (Table 2).Seven studies were conducted in Canada.
Over half (56%) of all articles reported data for specific age groups with the most common being adults/young adults (33%).
Eighteen articles (5%) reported factors as part of an ESBLproducing Enterobacteriaceae outbreak (all in hospital settings).
For most studies (53%), data were reported from secondary data sources (e.g., databases, medical charts), with 20% from primary data sources (e.g., questionnaires, interviews) and 7% from both primary and secondary data sources.For 20% of the studies, it was unclear how the data were obtained (Table 2).
Other articles sought to report different combinations of Enterobacteriaceae species (e.g., Klebsiella spp.and E. coli) (Table 2).
Most articles were performed in hospital settings (86%), followed by non-hospital healthcare settings (7%), community settings (6%), and residential care facilities (4%).Eleven of these articles were sampled from multiple of these different sample settings (Table 2).Overall, the highest number of articles identified for each factor theme were those that had performed their study in hospital settings, except for the community theme (Figure 2).
Articles reported factors for 1) infection (52%), 2) colonization/ carriage (33%), and 3) colonization/carriage/infection (13%) (Table 3).Factors potentially associated with ESBL-producing Enterobacteriaceae infections were reported in over half of the articles (52%) (mostly bloodstream infections or urinary tract infections).More articles identified factors for infection than colonization/carriage (Figure 3), especially for the factor themes antimicrobial use, demographics, comorbidities/symptoms and health care.Colonization/carriage was reported in a third of the articles (33%), with most focused on gastrointestinal carriage.Animal, community, food and consumption, household, occupation and travel themes were more frequently reported for colonization/carriage (Figure 3).For eight articles (2%) it was unclear whether the study was reporting colonization/carriage or infection.Acute pyelonephritis 5 1

Gastrointestinal and nasal 2 1
Gastrointestinal and vaginal 2 1 Gastrointestinal, vaginal and nasopharyngeal 1 1 Gastrointestinal, nasal and navel 1 1 Gastrointestinal, nasal, oropharyngeal and urine 1 1 Gastrointestinal, nasal and throat 1 1 Non-specific isolation 7 2 Urinary isolation 1 1 a Four articles had two extended-spectrum beta-lactamase-producing Enterobacteriaceae outcome evaluations  The same culture with any other bacteria than the compared ESBL-positive Enterobacteriaceae strain (i.e., cultures could be negative or positive for any bacteria blood culture except ESBL-producing E. coli) Abbreviations: E. coli, Escherichia coli; ESBL, extended-spectrum beta-lactamase; UTI, urinary tract infection a Twenty articles provided two comparisons b Control groups were non-ESBL producers but may have other resistance-susceptibility profiles c Control groups were negative for the same ESBL-producing Enterobacteriaceae colonization/carriage; however, the controls were positive or negative for the presence of other Enterobacteriaceae or non-Enterobacteriaceae cultures d Control groups were negative for the same ESBL-producing Enterobacteriaceae infection; however, the controls were positive or negative for the presence of other Enterobacteriaceae or non-Enterobacteriaceae cultures e Control groups were negative for the same ESBL-producing Enterobacteriaceae colonization/carriage and/or infection.However, the presence of other Enterobacteriaceae or non-Enterobacteriaceae cultures were not explicitly reported f Control group explicitly reported Enterobacteriaceae in combination with non-Enterobacteriaceae families (e.g., E. coli, Klebsiella spp., and Pseudomonas spp.) g Control group did not explicitly report bacterial species within the study; therefore, it was unclear whether the control group included only Enterobacteriaceae cultures or whether non-Enterobacteriaceae species cultures were included

Discussion
In this scoping review, we identified 381 articles reporting factors for ESBL-producing Enterobacteriaceae.Most of the included articles were published in the last 10 years, likely corresponding to the urgency to understand the growing rates of human acquisition of ESBL-producing Enterobacteriaceae and the exponential growth of scientific publications generally (8,(37)(38)(39).
It is noteworthy that most articles focused on factors related to antimicrobial use, comorbidities/symptoms, demographics and health care, and that only a small proportion of identified articles reported factors associated with animal contact, community, and food and consumption; mainly related to colonization/ carriage of ESBL-producing Enterobacteriaceae.Although there were fewer articles that reported these themes, they may provide important information as previous articles have suggested that animal contact, food consumption and household or community transmission may play a role in ESBL-producing Enterobacteriaceae exposure (11,17,(40)(41)(42).It is unclear whether the individual factors that were most frequently reported in these articles were in fact more often associated with ESBL-producing Enterobacteriaceae (i.e., had larger measures of association), whether they had been evaluated and reported more frequently than others, or whether studies evaluating these factors were better funded.
Study setting may be an explanation for the larger number of articles on antimicrobial use, comorbidities/symptoms, demographics and health care factors reported.Most articles were conducted in hospital settings and over half of the articles used secondary sources of information (e.g., medical records or databases).This setting and source combination may have been selected on account of the relative ease of accessibility to the data.Factors associated with resistant infections in hospitals are major concerns, and therefore are an important area of research.Although some factors reported from hospital settings may be connected to those in the community settings (e.g., taking medication), factors reported from hospital settings may not be representative of factors from community settings (e.g., populations, comorbidities, varying activities).Thus, the results from studies conducted in hospital settings are not generalizable to other settings.
This review identified studies where the subjects were sampled from countries with a very high human development index (33) as we were interested in factors relevant to the Canadian context.Most studies were conducted in the European Region (n=235), followed by the Western Pacific (n=78), the Americas (n=60) and the Eastern Mediterranean (n=8).Only seven studies were performed in Canada; however, a large body of literature was collected that can be used to understand the existing knowledge of factors associated with acquiring ESBL-producing Enterobacteriaceae in similar populations.Although these countries have similarities, differences in policies and practices may limit the generalizability of the data specifically to Canada.Work is currently underway to better describe the factors identified in these articles.This will provide the number of factors reported per study and quantitative data reported for these factors (i.e., the strength and direction of association between the factor and ESBL-producing Enterobacteriaceae). Further, factors from this review will be used to populate models within the iAM.AMR project (15,28) to improve our understanding of the pathways of human exposure to ESBLproducing Enterobacteriaceae.This information will help to inform which human characteristics, behaviours and actions impact the probability of becoming colonized or infected with ESBL-producing Enterobacteriaceae and to identify which factors to prioritize for interventions.This information will be valuable for understanding how to advise Canadians about mitigating their probability of acquiring resistant bacteria and reducing the negative health impacts associated with infection.

Limitations
Articles were identified from select online databases, omitting research from grey literature.This may have introduced a publishing bias, as findings that were not disseminated through peer-reviewed publications were not reviewed for inclusion (e.g., theses and dissertations, government reports) and articles

SCOPING REVIEW
with null, negative or inconclusive findings are less likely to be published (46).Language bias was a consideration as the review was constrained to English-language articles; however, the impact of this bias was likely negligible as approximately 98% of science publications are written in English (47,48).
Another limitation included single reviewer data extraction on account of resource limitations.Multiple individuals extracting study data reduces errors and misclassification bias (49).To mitigate these types of errors and to identify errors in data extraction, the authors were involved in both data collection development and analysis.
Lastly, the grouping of factors into themes evolved during data extraction.Grouping factors into themes was challenging because of differences in terminology used, the populations studied, and definitions applied.Combining data from different studies was onerous due to heterogeneity of the study data (e.g., same variable measured on different scales, missing data) (50).Terms, including carriage and colonization, were not standardized across studies and were used interchangeably; therefore, some data had to be combined (e.g., colonization and/or carriage) or captured as "unclear."

Conclusion
This review synthesized evidence from a large collection of articles reporting factors associated with ESBL-producing Enterobacteriaceae colonization, carriage and/or infections in humans within very high human development index countries.Factors were reported in many different settings, age groups and organisms, and using different outcome comparison groups.This variability between studies highlighted the need for transparent or, where possible, harmonized reporting of methods to allow for appropriate interpretations and comparisons between the factors reported.Overall, studies conducted in hospital settings predominated and the most common factor themes reported were antimicrobial use, comorbidities/symptoms, demographics and health care.Articles reporting animal contact, food consumption/practices and activities in the community were not as numerous and thus limited information about these factors were identified.There is a need for more studies examining factors associated with ESBL-producing Enterobacteriaceae in the community, which have been identified as being of concern (6,8).
This scoping review synthesized knowledge about potential sources and activities that affect the risk of human exposure to ESBL-producing Enterobacteriaceae.Factor themes identified spanned human, animal and environmental contexts and settings support the need for a diversity of perspectives and a multisectoral approach to AMR.The results of this article will help guide recommendations to reduce the risk of acquiring ESBL-producing Enterobacteriaceae for Canadians, as well as other similar countries, while considering numerous sources of exposure in various settings.These results will also guide future research for activities and in settings that are understudied.

Figure 1 :
Figure 1: Flow diagram of the interface search, screening process and included articles to identify articles reporting risk factors for extended-spectrum beta-lactamase-producing Enterobacteriaceae in humans a

Figure 2 :Figure 3 :
Figure 2: Factor themes reported in the articles by study sample setting for colonization/carriage and/ or infection with extended-spectrum beta-lactamaseproducing Enterobacteriaceae a

1 1 1
colonization/carriage and/or nonspecific infection 38 10 Urinary colonization/carriage and/or urinary tract infection 6 Gastrointestinal colonization/carriage and/or nonspecific infection 5 Respiratory colonization/carriage and/or infection 1 Urinary colonization/carriage and/or urinary tract infection,

Table 1 :
Description of factors represented by the factor themes for colonization, carriage and/or infection with extended-spectrum beta-lactamase-producing Enterobacteriaceae reported by the articles included in this review Abbreviations: ESBL, extended-spectrum beta-lactamase; ICU, intensive care unit

Table 1 :
Description of factors represented by the factor themes for colonization, carriage and/or infection with extended-spectrum beta-lactamase-producing Enterobacteriaceae reported by the articles included in this review (continued)

Table 2 :
Study characteristics of the included articles

Table 3 :
Description of reported extended-spectrum beta-lactamase-producing Enterobacteriaceae outcomes reported in the included articles

Table 3 :
Description of reported extended-spectrum beta-lactamase-producing Enterobacteriaceae outcomes reported in the included articles (continued)SCOPING REVIEWMany comparison groups were reported (Table4).The most common was an ESBL-positive Enterobacteriaceae culture compared with an ESBL-negative Enterobacteriaceae culture (n=171).Twenty articles reported two comparator groups (e.g., case-case-control studies).

Table 4 :
Reporting of outcome comparisons among articles for colonization/carriage and/or infection with extended-spectrum beta-lactamase-producing Enterobacteriaceae in the included articles